Transmission Loss Allocation in Pool-Bilateral Environment Using Artificial Neural Networks

Restructuring of power systems have introduced new challenges for operators, Suppliers and customer. In the new environment, special attention should be paid to losses and allocating it to market participants. Various strategies for loss allocation have been introduced in the past decade. Most of these methods rely on time consuming calculations. On the other hand, calculation time of loss allocation for a pool system is not crucial. The calculation time together with the accuracy of the method may become very important for bilateral contracts specially those taking place in power exchange. In this paper, a new approach for accurate estimation of losses allocated to various transactions is proposed. The method is based on utilization of a multilayer perceptron neural network. To investigate the accuracy of the method, it has been applied to IEEE 14-bus test system. The estimated values have been compared with those obtained through direct calculation. The results confirm the accuracy as well as the speed of the proposed algorithm.